MR. WHITEHURST: My remarks will be briefer than Lee's, and I think probably focus more on selfish and pragmatic concerns.

We have a series of problems to solve at the Institute of Education Sciences. The problems are largely practical problems.

They are issues and questions and concerns raised by policy makers and practitioners. They would like answers to questions like what sorts of math curricula work best for what kids under what circumstances.

If you are going to do professional development of teachers, what should the content be? How long should it last? Who should be in it? What should the measures of outcome be?

If you are going to spend 21 percent of the national budget in K-12 education on children in special education, how is that money being invested, and how can we invest it more wisely, and so forth and so on.

We also have a number of questions to address and answer in terms of evaluation of federal programs. Does Even Start work? If it doesn't, how could it be made to be more effective?

Do the after school programs work as they are currently funded? If not, how do they need to be changed?

We have a huge statistics enterprise that has many challenges in it.

These problems require well-trained scientists in order to provide the tools and solutions that are needed, and we have them in insufficient supply.

How do I know that? I know, from the applications that we received the last two years that we don't get enough strong applications.

The strong applications tend to be from very senior people in the field. If you look at the distribution of applicants across age, you don't find what is typically the case in other scientific areas I am familiar with, and that is a burst of activities by people in their first 10 years of career.

There seems to be a gap there, in terms of what we are seeing in terms of application for federal funding at IES.

I have looked at educational psychology. I know that, in the 2001, there were 16 educational psychologists who received PhDs in the previous year who were engaged in educational research the year thereafter. I am sure perhaps 15 of those were hired by ETS, and one went out in a university.

Perhaps I have missed a very careful accounting of the supply in the previous presentations here, but my sense, from the information that I have available is that the supply is not nearly sufficient to the demand for researchers.

The problems that I have described and have addressed in other forums are problems that require researchers who have a constellation of skills and abilities involving strong quantitative skills, strong methodological skills, and concomitant logical abilities, in order to test ideas and programs, to ask whether something works, to think about the evidence that might disconfirm a hypothesis and to move the field forward.

What would it take to educate students who have these skills in greater numbers? Here I have some insights that came from my long service as an academic at the State University of New York at Stony Brook, where I spent my last 15 years in the doctoral program in clinical psychology.

Clinical psychology actually has some interesting parallels to education, in that virtually everybody trained at the doctoral level goes into practice, not into research.

There is a sense of lower prestige for the practitioners, particularly those that get the Psy.D,. the doctor of psychology degree, versus the Ph.D.

How do you train researchers-in-training to address the critical problems in clinical psychology, without having them divert into careers in practice, and how do you provide a training program that will produce large numbers of researchers?

We actually had remarkable success at Stony Brook. The program generated more students who entered research careers than any other program in the country, and did so consistently over at least 20 years.

I think there are lessons that I learned as part of that experience that could be transferred to the training of students who are going to do education research.

Let me list some of those thoughts for you, for whatever they may be worth. What are the ingredients of a successful doctoral training program in research.

First is a productive faculty. Research training programs are largely mentorships. In the absence of a strong and productive faculty, it is very unlikely that any amalgam of course work will generate large numbers of students who are well prepared to enter the research profession.

Those faculty have to be not simply productive. They also need to share a world view about what the enterprise is about, what the questions are, and how you approach them, not a narrow set of shared assumptions, but an assumption about the nature of the science in which everybody is engaged.

You can permit an outlier or two, but if there is a real diffusion of the view, it is unlikely that you will have a cohesive faculty and, in the absence of a cohesive faculty and shared goals for training, it is difficult for me to see how you will obtain or generate a focused and effective training program.

It may be necessary to isolate those faculty, if they are a subset of a larger group of faculty that does not share those goals, to isolate them in a particular program.

It may be that, in many circumstances, in order to achieve a necessary mass of faculty who share goals and assumptions, you may need to cross disciplinary boundaries.

So, it may well be the case that the training of scientists in education need not be located exclusively in the discipline of education, but can include faculty members in education with faculty members in other departments who are focused on education problems.

The second ingredient for this pie we hope to bake here is selective admissions, and I am talking about selective admissions not only in the traditional sort of selection, where you look at GRE scores, but selective admissions in terms of career goals.

At Stony Brook, to be admitted into the graduate training program in clinical psychology, you had to profess an interest in having a research career and not be easily talked out of it in an interview.

The program was clearly advertised as exclusively for research training, come here, this is what you get, if this is not what you want, go some place else. We do not train practitioners here.

If you select according to that goal, you are more likely to get people who are actually interested in research careers. Absent that motivation, you will get an occasional researcher, but not researchers in numbers.

A third suggestion -- and this aligns with one of Lee's comments -- you need to support these students. They need a full stipend for four years. They need to be in residence full time. They should not work more than five hours a week on activities that are unrelated to their training as a researcher.

So, if they are going to be a teaching assistant, five hours a week is just fine. If they are going to go off and do something in terms of clinical practice or if they are going to work in school, five hours or so is fine.

Once it escalates beyond that, students are seduced into the early reinforcers of helping. Those reinforcers tend to move many students out of the career of research, which takes a considerable delay of gratification.

Fourth, build a culture around these students. Have them think that they are the best and the brightest, that they are doing the most important thing that could be done in terms of graduate training. This creates cocky, irritating students, but the prognosis for them, long term, I think, is positive.

Include in their training -- this is point number five -- multiple, progressive, research-related products. There should be a first year product, there should be a second year product, there should be a third year product, there should be a fourth year product.

That experience, and the expectation that publications will flow from those experiences in some number, generates a set of skills and experiences that prepare students for the career of research.

I am still surprised at the number of applications we get at the Institute of Education Sciences for people who want to come on board and be part of what we are doing as a research agency, and who have received a PhD, and yet, have no publications to show for it.

Starting to conduct serious research and publishing it is not something you do after you have got your Ph.D. if you want a career in research.

Frequently, at Stony Brook, our students would generate five, six, seven or more publications prior to receiving a PhD. They were doing what they were going to be expected to do as assistant professors, and they typically succeeded in their first jobs.

Part of the multiple progressive research products at my former institution was to require mentorship by at least two people.

One of the required research products had to be generated in somebody else's lab – somebody other than the primary advisor. That generates a different set of research skills and different approaches, which often provide the ability to have insight that is less likely to occur if you are narrowly focused and have had only one set of experiences.

The sixth ingredient was a common methods course that was rigorous, lasted a year, was very demanding, and required a variety of hands-on research projects using an array of research tools.

Students would go through multivariate approaches, through experimental approaches, through observational approaches and, in each case, would have to work with existing data sets, or collect data, analyze it, turn in results, and have those results critiqued by sophisticated, advanced graduate students.

It provided a set of tools that should be available in the tool kit of any researcher who anticipates a career that is going to last for a while, and that requires grounding in a variety of techniques.

The seventh point was a commitment to placement. We worked hard to get our students in academic positions, and to provide them early career support, so that they could succeed there.

The eighth point -- and here I differ with Lee -- is that, to start such a program, you need a different cutting edge focus.

There has to be a reason for highly qualified students to want to apply, to think of the institution as one that is offering something they are unlikely to get anyplace else, and to be attracted to it.

The enterprise I have in mind is explicitly elitist. You don't need 100 institutions providing doctoral level training in research. You need seven, eight, nine or ten.

They have to be the sort of institutions that advertise themselves, that are advertising something real, and that generate enthusiastic applications from students who are interested in what that institution is able to offer to them.

The ninth point is the need for some national organization around which researchers with the combination of skills and attitudes I have described can coalesce. I am not sure that exists in education research.

The AERA is a wonderful organization. It is not, I think, substantially different in its relationship to the point I am making than the American Psychological Association is for psychologists.

The American Psychological Association is not the place that research psychologists identify wiath and come to meetings with respect to their primary interests. They come occasionally, but it is not their organization.

I think if we are going to have a serious committed research community in education, there needs to be some other organization at which they feel particularly at home.

The tenth point is: we need funding for all of this. It cannot be created out of whole cloth. Schools of education are particularly challenged. They are often seen as cash cows on their campuses. They are expected to generate the revenue to keep the institution afloat.

That is often incompatible with a very significant investment of faculty time and stipend time in the sort of graduate training I am talking about.

I have been shocked to talk to colleagues in schools of education and find that they are carrying 20 doctoral advisees.

That is quite remarkable. How can you provide mentorship for such a large number of students? Three or four at a time should be quite enough. That needs substantial funding.

What we have done, at the Institute of Education Sciences, is to try to get that funding. We had it in last year's budget proposal. It is in this year's budget proposal. We are still waiting with fingers crossed. It is roughly $18 million to support new interdisciplinary pre- and post-doctoral training programs in education sciences, programs that would be competitive, from universities that can put together proposals from stellar faculty across disciplinary units, to propose how they would train students to enter a career in education research.

I think $18 million is a good starting point. I think we certainly need more than that to support the type of enterprise that I am talking about here.

I am very optimistic that we are going to see not only increased funding for education research -- and that is another part of this as well. You need graduate students who can look at a career in education research and think to themselves, rationally, there is a career to be made here, there is stable funding. I can see myself going into a university and getting tenure for doing this research. I can see myself building a career that would last 25 or 30 years and being assured that, if I am at the top of my field, I can continue to get funding. We need that funding as well as graduate student training, to turn this around.

I think that funding will be there, and I am very optimistic that we will see substantial progressive changes in the nature of the education of graduate students who are going to enter education research careers in the next decade or so. Certainly we need it. Thank you very much.

[Applause.]